EP0532988B1 - Process for preparing random copolyesters - Google Patents

Description

The invention relates to a process for producing statistical copolyesters with a low comonomer content from homopolyester and copolyesters with a higher comonomer content, in particular from polyethylene terephthalate and isophthalic acid and / or 1,4-bis (hydroxymethyl) cyclohexane-containing co-polyethylene terephthalate.

Statistical copolyesters are usually prepared by subjecting the mixture of the monomers to esterification or transesterification with subsequent polycondensation. One or more comonomers can also be added to the precondensation. Large-scale technical plants usually comprise several reactors connected in series, including auxiliary units such as glycol recovery and cycles, and require a high investment volume. A producer who wants to produce large quantities of homopolyesters, for example polyethylene terephthalate for fibers, smaller quantities of copolyester, for example modified polyethylene terephthalate for bottles, is faced with the economic problem of either building an additional production line for the copolyester or notable quantities of not - Accept homopolyester / copolyester / homopolyester during the conversion phases.

It is also known to mix 1,4-bis (hydroxymethyl) cyclohexane-containing copolyethylene terephthalate with approximately the same amount of homo-polyethylene terephthalate in the melt and, if appropriate, then to subject this mixture to an additional heat treatment after conversion into powder form (US Pat. A 3 651 172). However, the polyesters obtained in this way are block copolyesters which differ significantly in their properties from statistical copolyesters and are consequently not suitable for the same applications.

The object of the present invention is therefore to find a process for producing statistical copolyesters which is associated with low investment costs and in which practically no non-type-appropriate product is obtained. In particular, in the case of existing production of large amounts of polyethylene terephthalate, for example for fibers, the task relates to ensuring the most economical production of smaller amounts of co-polyethylene terephthalate for hollow bodies and bottles.

This object is achieved in accordance with the disclosure of the claims.

The polyester (A) to be used is polyethylene terephthalate of customary quality and composition, as used, for example, for fiber production. The intrinsic viscosity is 0.5 to 0.8 dl / g, preferably 0.55 to 0.65 dl / g, and the diethylene glycol content is 0.6 to 2.0% by weight. Coming from the polycondensation catalyst, it normally contains 50-300 ppm Sb. Depending on the local conditions, it is assumed that the granulate has melted again or a partial stream of the melt emerging from a polycondensation reactor is passed directly to the mixing unit to be used according to the invention.

The polyester (B) is a prepolymer with an intrinsic viscosity of 0.1 to 0.6 dl / g, preferably 0.15 to 0.30 dl / g, which contains at least 10% by weight of isophthalic acid and / or 1,4-bis (hydroxymethyl) cyclohexane, the difference being 100 mol% of dicarboxylic acids or diols of terephthalic acid and ethylene glycol, which can be replaced by up to 50 mol% of diethylene glycol . The polyester (B) thus contains two to five components, selected from isophthalic acid, terephthalic acid, 1,4-bis (hydroxymethyl) cyclohexane, ethylene glycol and diethylene glycol, the sum of the isophthalic acid units and the 1,4-bis- ( hydroxymethyl) cyclohexane units must be at least 10% by weight, based on the polyester (B), and the proportion of diethylene glycol units in mol% being less than or equal to the proportion of ethylene glycol units. The term polyester (B) here also encompasses the homopolyesters, polyethylene isophthalate, poly (cyclohexane-1,4-dimethylene) terephthalate and poly (cyclohexane-1,4-dimethylene) isophthalate. The statistical distribution in the final copolyester is favored by the relatively low viscosity or short-chain nature of the polymer molecules.

Melted prepolymer granules can be used. However, the prepolymer is preferably produced directly in a simple esterification or transesterification and precondensation system comprising one or a maximum of two reactors. The capital expenditure for such a prepolymer plant is only a fraction of the costs required for a complete polyester plant. A part of the reactors and ancillary systems is omitted, and the remaining equipment parts are less expensive.

For the prepolymer production, terephthalic acid or its ester and / or isophthalic acid or its ester as well as ethylene glycol, which can be replaced by up to 50 mol% by diethylene glycol, and / or 1,4-bis (hydroxymethyl) cyclohexane at about normal pressure and usual temperatures in the range of 230 to 290 ° C. In view of the further treatment according to the invention and a distribution as homogeneous as possible in the polymer, the addition of 10 to 1000 ppm of metal of a conventional polycondensation catalyst, preferably an antimony, titanium or germanium compound, such as, for example, antimony triacetate, glycolate or trioxide, is recommended. Titanium tetra (oxyalkylate), germanium dioxide or hydrogen phosphite. The catalyst concentration in the prepolymer (B) is preferably 100 to 200%, particularly preferably about 150% of the amount of polycondensation catalyst present in the polyester (A). The prepolymer can be produced continuously or batchwise. In the latter case, an intermediate storage vessel for the melt or for the granulate must be provided.

Prepolymer melt (B) and polyester melt (A) are fed to a mixing unit and react partially, with homogenization in the molten state, for 1 to 50 minutes, preferably 5 to 15 minutes. The quantitative ratio of the two melt streams results purely mathematically from the comonomer content of the prepolymer (B) and the desired comonomer content of the statistical copolyester (C) to be produced. The entire polyester (B) is preferably first mixed with 10 to 50% by weight of the polyester (A) in a mixing extruder and the extrudate is then mixed in a static mixer with the remaining 90 to 50% of the polyester (A). This enables less equipment outlay, since a smaller extruder is sufficient. The mixing time, defined as the time from the meeting of the melts of the polyesters (A) and (B) until the mixture enters the subsequent granulator, is determined by the homogenizability of the commercially available mixing units used and the volume of these apparatuses and their connecting lines.

Due to its melting point, which differs only slightly from that of the polymer (A), the melt mixture entering the granulator can be regarded primarily as a physical polymer mixture. The conversion into a statistical copolymer is reserved for the subsequent solid phase treatment.

The statistical copolyester to be produced according to the invention is primarily intended for processing into hollow bodies and bottles, that is to say for areas of application in which color and gloss play a special role. In this case, the prepolymer melt or the mixture of both melts becomes a blue toner at the beginning of the mixing phase, based on the statistical copolyester (C), consisting of a maximum of 5 ppm of a polyester-soluble mixture in a ratio of 3: 1 to 1: 3. preferably 1: 1, a red and a blue dye, preferably of Estofilrot S-RBL and Estofilblau S-GFP from Sandoz, or from 10 to 60 ppm Co plus 10 to 60 ppm P in the form of their compounds.

Furthermore, especially if processing is to be carried out by extrusion blowing and a higher melt strength is desired, 100 to 3000 ppm, based on copolyester (C), of a trifunctional or tetrafunctional comonomer such as trimethylolethane, trimethylolpropane, glycerol or pentaerythritol can be used at the beginning of the mixing phase , are added.

The melt mixture is granulated in a known manner, attention being paid to the size of the chips, corresponding to 0.5-5.0 g / 100 chips, preferably 1.0-2.0 g / 100 chips. Too small chips, powder for example, hinder the mass transfer in the solid phase and thus the formation of a statistical distribution within the polymer chains.

The granules are then initially crystallized under inert gas, for example nitrogen, at a temperature T 1 above 150 ° C. and below the melting temperature for 0.5 to 3.0 hours and then under inert gas or vacuum at a temperature T 2 above 150 ° C. and below or equal to T₁ post-condensed until the desired intrinsic viscosity is reached. This normally takes 1 to 20 hours. Granules which contain prepolymer with a higher addition of polycondensation catalyst require shorter post-condensation times than those with prepolymer with a low addition of catalyst.

Surprisingly, it was found that the granules treated in this way have a melting temperature, measured by means of DSC (differential scanning calorimetry), which is not more than 2 ° C. higher than the measured melting temperature of statistical copolyesters of the same summary composition. The other properties of the copolyester produced according to the invention are also identical to those of statistical copolyesters which were obtained by direct polycondensation of the monomers.

wobei für DEG, IPA und CHDM die Konzentration in Gew.-% des jeweiligen Comonomers im Copolyester einzusetzen ist.From preliminary tests, it was determined that the melting temperature T _M, measured by DSC, of statistical copolyesters produced by direct polycondensation, the comonomers diethylene glycol (DEG), isophthalic acid (IPA) and / or 1,4-bis (hydroxymethyl) cyclohexane (CHDM) contain up to a total concentration of 5 wt .-%, can be determined according to the following formula: $${\text{T}}_{\text{M}}\text{(° C) = 261 - (DEG. 4.20) - (IPA. 3.04) - (CHDM. 4.66)}$$

the concentration in wt.% of the respective comonomer in the copolyester being used for DEG, IPA and CHDM.

In contrast to the described prior art, statistical copolyesters are obtained according to the invention, not block copolyesters. The reasons for this are not fully understood, but it is certain that if the melt mixing time claimed is adhered to, statistical copolymers are only formed if the polyester (B) is relatively low molecular weight and the chips do not fall below a minimum size. When using a polyester (B) with an intrinsic viscosity of more than 0.6 dl / g, statistical copolymers are only obtained if the melt phase treatment is significantly extended until a predominantly statistical distribution in the melt is reached. The post-condensation is otherwise not sufficient for the necessary Mokekül compensation. With such a long melt phase treatment, it is difficult to avoid discoloration of the polymer and gel formation.

Statistical copolyesters (C) with a somewhat higher comonomer content than 5% by weight, for example up to about 10% by weight, can be prepared analogously by the process according to the invention.

The following examples support the invention without restricting it.

The intrinsic viscosity (I.V.) was determined on a solution of 0.5 g polyester in 100 ml of a mixture of phenol and 1,2-dichlorobenzene (3: 2) at 25 ° C.

Diethylene glycol (DEG), isophthalic acid (IPA) and 1,4-bis (hydroxymethyl) cyclohexane (CHDM) were determined by gas chromatography after pressure-methanolysis of the polyester.

The COOH end groups were determined by titration.

The melting temperature (T _M ), the glass transition temperature (T _G ) and the crystallization temperature (T _K ) were determined by means of DSC, the sample being briefly heated to 300 ° C. before the actual measurement and immediately quenched.

Comparative examples (examples 1 and 2): example 1

8000 g terephthalic acid
272 g isophthalic acid
3422 g ethylene glycol
70 g diethylene glycol
180 ppm antimony as antimony triacetate (based on polyester)
1.5 ppm Estofilrot S-RBL (based on polyester)
1.5 ppm Estofilblau S-GFP (based on polyester)
20 ppm phosphorus as H₃PO₄ (based on polyester)
were stirred in a mixer to form a stable paste, which was then metered at a constant rate over a period of 120 min into an esterification vessel equipped with a stirrer, column and jacket heater, in which 50% of the esterification product of the same composition were from the previous batch. The temperature was maintained at 261 ° C during the feed time.

After the paste feed had been completed, the temperature was raised to 270 ° C. over a period of 30 minutes under normal pressure and then the pressure was reduced to 50 mbar over a period of 30 minutes while increasing the product temperature to 278 ° C. The esterification product thus obtained was then transferred to a polycondensation reactor and polycondensed for 160 minutes at a temperature of 281 ° C. and a pressure of 0.70 mbar. After the polycondensation had been terminated, the contents of the reactor were granulated over a period of 15 min. The average weight of the granules was 15 mg. The polyester granules thus produced were then crystallized under nitrogen for 60 minutes at 215 ° C. and then polycondensed at 210 ° C. for 10 hours (SSP). Table 1 summarizes the analytical parameters of the intermediate and end products.

Example 2

The polyester is produced as described in Example 1, but without the use of isophthalic acid and diethylene glycol, 1,4-bis (hydroxymethyl) cyclohexane (CHDM) being used instead of these comonomers according to the following recipe:
8272 g terephthalic acid
3422 g ethylene glycol
161 g CHDM
Antimony and blue toner as in example 1.

The analytical parameters are shown in Table 1. Table 1 Analysis values of the comparison products example 1 Example 2 amorphous according to SSP amorphous according to SSP IV (dl / g) 0.642 0.785 0.625 0.791 COOH (meq / g) 20th 14 19th 14 DEG (% by weight) 1.21 1.21 0.57 0.57 IPA (% by weight) 2.82 2.82 0 0 CHDM (% by weight) 0 0 1.57 1.57 T _M (° C) 247.0 246.5 251.4 250.4 T _K (° C) 157.0 157.0 142.3 145.3 T _G (° C) 77.4 77.5 79.3 78.4 T _M calculated (° C) 247.3 251.2

Preparation of the polyester (A): Example 3

8272 g of terephthalic acid were reacted with 3422 g of ethylene glycol as in Example 1 to give polyester, 180 ppm of antimony being used as the antimony triacetate as catalyst. The polyester thus obtained had an intrinsic viscosity of 0.63 dl / g, a DEG content of 0.87% by weight and 28 meq / g COOH end groups. The melting temperature, glass transition temperature and crystallization temperature were determined by DSC as follows:
T _M 256 ° C; T _G 79.8 ° C; T _K 148 ° C.

Preparation of the polyesters (B) (Examples 4 to 7): Example 4 Isophthalic acid and ethylene glycol:

8272 g of isophthalic acid were esterified together with 3422 g of ethylene glycol and 180 ppm of Sb as Sb₂O₃ in a stirred reactor at an internal pressure of 1.5 bar and a temperature of 261 ° C with elimination of water. After completion of the esterification, the pressure was released to normal pressure and then polycondensed while reducing the pressure to 50 mbar at a temperature of 270 ° C. until an intrinsic viscosity of 0.15 dl / g was measured. After breaking the vacuum with nitrogen, the precondensate was cooled to 150 ° C., mixed with a mixture of the dyes mentioned in Example 1 in a ratio of 1: 1 in an amount of 3 ppm, based on copolyester (C), and with further cooling granulated.

Example 5 15% by weight IPA in polyester (B):

6825 g of terephthalic acid, 1447 g of isophthalic acid and 3422 g of ethylene glycol were, as in Example 4, processed to polyester with an intrinsic viscosity of 0.60 dl / g, instead of 180 ppm of antimony, 300 ppm of antimony, based on the finished copolyester ( B) were used.

Example 6 18% by weight CHDM in polyester (B):

7500 g of terephthalic acid, 1737 g of 1,4-bis (hydroxymethyl) cyclohexane and 2400 g of ethylene glycol were, as in Example 6, processed into a precondensate with an intrinsic viscosity of 0.30 dl / g.

Example 7 35% by weight CHDM in polyester (B):

6890 g of terephthalic acid, 3600 g of 1,4-bis (hydroxymethyl) cyclohexane and 1250 g of ethylene glycol were, as in Example 6, processed into a precondensate with an intrinsic viscosity of 0.40 dl / g. In contrast to the recipe mentioned in Example 6, no dye was added, but cobalt as cobalt acetate was additionally added at the beginning of the esterification in such an amount that 40 ppm of Co are present in the end product (C). In addition, 40 ppm of phosphorus was added as triethyl phosphate, based on the end product (C), shortly before the end of the reaction.

Preparation of the statistical copolyesters (C) (Examples 8 and 9): Example 8

Mixtures of polyester (A) from Example 3 and polyester (B) from Examples 4 to 7 were carefully dried and melted in an extruder to which a mixing section consisting of a melt line equipped with static mixers was melted and after a residence time of 10 min granulated. The granules had an average weight of the individual grains of 14-17 mg.

The granules were crystallized under nitrogen at 215 ° C. for 1 hour and then subjected to solid polycondensation (SSP) at 210 ° C. for 7 hours. The amount of N₂ delivered per hour by the SSP reactor was 1 kg of N₂ per 1 kg of polyester.

All polyester mixtures were examined for their thermal DSC data before and after the SSP. The results are summarized in Table 2.

Example 9

As in Example 8, the polyester (A) from Example 3 was processed with the polyesters (B) from Example 4 or 6 in the mixing section, with an additional 700 ppm, based on the end product (C), of pentaerythritol being added. All other steps were carried out as described in Example 8.

The results are summarized in Table 3. Table 2 Production of statistical copolyesters (C) Mixture of polyester (B) from example no. 4th 5 6 7 Quantity, parts by weight 1 2nd 2nd 1 + Polyester (A) Quantity, parts by weight 32 9 10th 17th after mixing: IV dl / g 0.565 0.653 0.590 0.610 COOH meq / g 29.0 28.5 28.0 28.6 T _M ° C 252.7 251.4 250.3 253.8 T _K ° C 134.5 138.0 146.1 141.2 T _G ° C 79.0 79.4 79.9 79.7 according to SSP: IV dl / g 0.735 0.855 0.751 0.780 COOH meq / g 25.0 24.0 23.0 23.0 T _M ° C 250.2 248.0 245.0 249.8 T _K ° C 143.8 146.0 159.0 151.9 T _G ° C 80.9 80.9 81.3 81.3 T _M calculates ° C 249.9 248.0 243.5 248.2 IPA% by weight 2.4 3.1 - - CHDM% by weight - - 2.9 1.9 DEG% by weight 0.90 0.85 0.93 0.92 Production of statistical copolyesters (C) with tetrafunctional comonomer Mixture of polyester (B) from example no. 4th 6 Quantity, parts by weight 1 2nd + Polyester (A) Quantity, parts by weight 32 10th + Pentaerythritol 700 ppm 700 ppm after mixing: IV dl / g 0.618 0.650 COOH meq / g 28.5 29.0 T _M ° C 252.3 250.1 T _K ° C 134.8 147.8 T _G ° C 79.4 80.1 according to SSP: IV dl / g 0.98 1.08 COOH meq / g 23.4 22.8 T _M ° C 250.0 245.3 T _K ° C 148.3 155.0 T _G ° C 81.1 81.1 T _M calculates ° C 249.9 243.8 IPA% by weight 2.4 - CHDM% by weight - 2.9 DEG% by weight 0.91 0.88 Pentaerythritol% by weight 0.07 0.07

Claims (8)

1. A process for preparing statistical copolyesters (C) containing, in addition to terephthalic acid units and ethylene glycol units, 0.5 to 2.0 wt.% of diethylene glycol units and 0.5 to 5.0 wt.% of isophthalic acid and/or 1,4-bis-(hydroxymethyl)-cyclohexane units, whose intrinsic viscosity measured in accordance with the description is 0.7 to 2.0 dl/g and whose melting temperature, measured by means of DSC, is not more than 2°C higher than the melting temperature of copolyesters of the same empirical composition obtained by direct polycondensation of the monomers, characterised in that

   - a polyester (A) with an intrinsic viscosity of 0.5 to 0.8 dl/g, consisting of polyethylene terephthalate with 0.6 to 2.0 wt.% of diethylene glycol units and

   - a polyester (B) with an intrinsic viscosity of 0.1 to 0.6 dl/g containing, with respect to polyester (B), 10 to 1000 ppm of Sb or Ti or Ge and at least 10 wt.% of isophthalic acid and/or 1,4-bis-(hydroxymethyl)-cyclohexane, wherein the difference to make 100 mol-% of each of the dicarboxylic acids or diols comprises terephthalic acid and ethylene glycol respectively, which can be replaced by up to 50 mol-% of diethylene glycol,

   - are mixed in the molten state for 1 to 50 min, are then granulated and the granular material with a chip-weight of 0.5 to 5.0 g/100 chips is crystallised under an inert gas at a temperature T₁ higher than 150°C and less than the melting temperature for 0.5 to 3 hours and then postcondensed under an inert gas or under a vacuum at a temperature T₂ higher than 150°C and less than or equal to T₁ until the desired intrinsic viscosity is achieved.
2. A process according to Claim 1, characterised in that all of polyester (B) and 10 to 50 wt.% of polyester (A) are fed to a compounding extruder and the extrudate is then mixed with the remaining 90 to 50 wt.% of polyester (A) in a statistical mixer.
3. A process according to Claim 1, characterised in that 0.5 to 5 ppm, with respect to statistical copolyester (C), of a polyester-soluble mixture of a red and a blue colorant in the ratio by amounts of 3 : 1 to 1 : 3 are added to polyester (B) before the mixing zone or to the mixture of polyesters (A) and (B) at the start of the mixing phase.
4. A process according to Claim 1, characterised in that 10 to 60 ppm of Co and 10 to 60 ppm of P, with respect to statistical copolyester (C), in the form of their compounds are added to polyester (B) before the mixing zone or to the mixture of polyesters (A) and (B) at the start of the mixing phase.
5. A process according to Claim 1, characterised in that 100 to 3000 ppm, with respect to statistical copolymer (C), of a tri or tetrafunctional comonomer are added to the mixture of polyesters (A) and (B) at the start of the mixing phase.
6. A process according to Claim 1, characterised in that polyester (A) contains an amount (m) in the range 50 to 300 ppm of Sb and polyester (B) contains an amount in the range (m) to (2m) of Sb or Ti or Ge.
7. A process according to Claim 1, characterised in that the intrinsic viscosity of statistical copolymer (C) is 0.7 to 0.9 dl/g, that of polyester (A) is 0.55 to 0.65 dl/g and that of polyester (B) is 0.15 to 0.30 dl/g.
8. A process according to Claim 1, characterised in that polyesters (A) and (B) are mixed for 5 to 15 min in the molten state and the chip-weight is 1.0 to 2.0 g/100 chips.